In 3GPP (3rd Generation Partnership Project), the cellular packet-switched communication systems HSPA (High Speed Packet Access) and LTE (Long Term Evolution) have been specified for radio transmission of data packets between user terminals and base stations in a cellular/mobile network. Transmissions from the base station to the user terminal is referred to as “downlink” and transmissions in the opposite direction is referred to as “uplink”. In the following description, “terminal” is used to generally represent any user equipment, commonly referred to as “UE” in the above systems, that is capable of wireless communication, e.g. with base stations in a cellular/mobile network.
There are two basic modes of operation available for wireless transmissions: FDD (Frequency Division Duplex) and TDD (Time Division Duplex). In FDD, downlink and uplink transmissions are made at separate frequency bands, such that packets can be transmitted in the downlink and uplink at the same time without mutual interference. In TDD, on the other hand, downlink and uplink transmissions are made on the same frequency band and must therefore be separated in time to avoid interference.
The TDD operation mode is flexible in that the duration of downlink and uplink transmissions can be configured depending on the traffic intensity in the respective downlink and uplink directions, thus allowing for connections with asymmetric transmission schemes. For downlink intensive connections, the downlink time period may thus be configured greater than the uplink time period, and vice versa for uplink intensive connections.
For LTE, a new physical layer is currently being standardized in 3GPP that is based on OFDM (Orthogonal Frequency Division Multiplexing) in the downlink and SC-FDMA (Single Carrier Frequency Division Multiple Access) in the uplink. The new physical layer shall support both FDD and TDD operation, and there should be a high degree of commonality between these two modes of operation. The SC-FDMA properties in the uplink require that any data transmitted from each terminal basically maintains single carrier properties.
The transmissions in both FDD and TDD operation are generally scheduled in radio frames, and each radio frame is typically divided into multiple sub-frames. In the following description, the term “sub-frame” is used to generally represent a predefined transmission time interval “TTI” in which a portion of information can be transmitted as a “data block”, although not limited to any particular standard or duration. A data packet can be sent in any number of sub-frames depending on the packet size and the sub-frame length. LTE prescribes that one data packet is typically accommodated in a single sub-frame. A sub-frame can generally contain one or more data blocks, in LTE also called “transport blocks”. Currently, LTE allows for two transport blocks per terminal in a single downlink sub-frame.
In LTE, the predefined radio frame is 10 ms (milliseconds), which is divided into ten predefined sub-frames of 1 ms each. In the FDD mode, where packets can be transmitted in the downlink and uplink simultaneously, there are 10 downlink sub-frames “DL” and 10 uplink sub-frames “UL” available during one radio frame on separate frequency bands F1 and F2, respectively, as illustrated schematically in FIG. 1a. In the TDD mode, there are in total 10 downlink and uplink sub-frames available during one radio frame, which can thus be transmitted only one at a time on a common frequency band F.
As mentioned above, downlink and uplink transmissions can be configured in TDD depending on the traffic demands in either direction. For example, the downlink/uplink allocation can be configured to 8 downlink sub-frames and 2 uplink sub-frames during one radio frame on the same frequency band F, as illustrated schematically in FIG. 1b. Another possible configuration could be 5 DL:5 UL sub-frames, and yet another configuration could be 2 DL:8 UL sub-frames. The alternation pattern of downlink/uplink sub-frames can also be configured optionally. For example, the downlink/uplink sub-frame pattern in FIG. 1b could be modified into 8 successive downlink sub-frames followed by 2 uplink sub-frames.
A single base station may transmit data packets in sub-frames on the downlink to one or more terminals, and the terminals transmit data packets in sub-frames on the uplink to the base station. The transmission in either direction is typically subjected to various disturbances, including propagation fading and interference from reflections and other transmissions, such that errors may have been introduced in the data packets when received. Thus, the channel between a base station and a terminal is often referred to as a “lossy” channel. Errors may also arise due to a poor receiver and/or antenna.
When receiving a packet with data in a sub-frame, the receiver in the terminal or in the base station is configured to check as to whether any errors are present in the received packet. A common method of detecting errors involves calculation of a check-sum or the like, which is well-known in the art. To enable correction of such errors, the data sending party must retransmit any erroneously received packet, unless some error correction mechanism can be applied successfully at the data receiving party. Therefore, the receiving party is typically obliged to send a feedback report to the data sending party for each received packet or sub-frame, indicating if the packet was basically received correctly, i.e. without errors, or not.
If the packet was received correctly, the data receiving party sends an acknowledgement “ACK”, and if the packet contained errors, it sends a negative acknowledgement “NACK”. Although the terms ACK and NACK are frequently used this description, any equivalent or similar messages may be used for feedback reports and the present invention is not limited in this respect. “Feedback report” is used in the following as a generic term for such ACK/NACK messages and their equivalents.
Both HSPA and LTE employ a HARQ (Hybrid Automatic Repeat ReQuest) protocol in their respective MAC (Medium Access Control) layers. The basic functionality of the processes defined in the HARQ protocol is to correct any erroneously received packets by means of retransmission based on the above-described feedback reporting mechanism. In this context, a feedback report is sometimes called “HARQ status report”.
For example, the data receiving party can simply discard an erroneously received packet. In more advanced solutions, the receiving party stores the signal representing the erroneously received packet in a buffer and combines this stored information with the retransmission. This is often referred to as “HARQ with soft combining” which can be used to increase the probability of correctly decoding the transmitted packet. In HARQ with soft combining, the pattern of coded bits in a particular packet may differ between transmission and retransmission, although they must obviously represent the same information.
The HARQ process is used to associate a potential retransmission to its original transmission in order to enable the soft combining at the data receiving party. When the receiving party has reported correct reception of data sent on a HARQ process, that data can be used to transmit new data. Consequently, before the reception of a HARQ status report from the receiving party, the data sending party does not know whether it should transmit new data or retransmit the “old data”. In the meantime, the sending party therefore “stops and waits” until the result of the transmission is reported. In order to still be able to utilize the link during these waiting periods, multiple parallel HARQ processes can be applied which allows for continuous transmission.
For example, when a packet is transmitted on the downlink, the receiving terminal checks for errors in the packet and sends a feedback report to the base station. If the base station then detects a NACK, it will retransmit the information in the packet. This mechanism can also be used for packets sent on the uplink. In LTE, the feedback required for HARQ with soft combining is conveyed by a single bit indicating either ACK or NACK. The timing relation between the packet transmission from the sending party and the feedback report transmission from the receiving party is typically used to indicate which packet the feedback report relates to.
In FDD, the number of available sub-frames is equal in the downlink and the uplink, as shown in FIG. 1a. Consequently, it is possible to send a feedback report for a data block received in one downlink sub-frame in a given uplink sub-frame according to a “one-to-one relation”, using a fixed time interval between reception and feedback. Thereby, the data sending party can derive which HARQ process a received feedback report refers to, based on which sub-frame the report was received in. In TDD, on the other hand, data blocks in multiple sub-frames may be received on the downlink before it is possible to send corresponding feedback reports, or ACK/NACKs, on the uplink, such as when the number of allocated downlink sub-frames is greater than the number of allocated uplink sub-frames.
In the allocation example of FIG. 2, there are 8 downlink sub-frames but only 2 uplink sub-frames available. Hence, feedback reports for the 8 downlink sub-frames must be transmitted in the 2 uplink sub-frames. Depending on how many users that have been scheduled in the downlink sub-frames, the number of feedback reports that need to be transmitted may increase by a factor 4. Furthermore, if a single terminal has been scheduled to receive data in all available downlink sub-frames, that terminal will need to transmit feedback reports for multiple data blocks received in a plurality of downlink sub-frames during a single uplink sub-frame. Still further, more than one data or transport block may be accommodated in a single received sub-frame, e.g. relating to one or more different sessions or media streams on a higher level, where each data block needs a separate feedback report such that the number of necessary feedback reports may increase even more.
In TDD, the above-described report mechanism with a fixed time interval cannot generally be used, since the feedback report for a received sub-frame cannot be transmitted a fixed time interval after receiving the sub-frame if the corresponding sub-frame is not available for transmission from the data receiving party. Consequently, the feedback report for data in that received sub-frame must be delayed at least to the first sub-frame available for transmission. Moreover, the data receiving party typically requires a certain delay after receiving a sub-frame, for processing the data therein and to determine if it was received correctly or not, before a feedback report can be sent for that sub-frame. For example, if the receiver needs a delay of at least 1 sub-frame for processing, a received sub-frame k cannot be reported until sub-frame k+2 or later.
A straightforward and obvious solution is to send a feedback report for a received sub-frame in the first available sub-frame after a minimum delay period needed for processing. Hence, if one or more sub-frames after the delay period are allocated for reception, the feedback report is further delayed until the first sub-frame available for transmission occurs. As a result, a plurality of feedback reports must typically be sent in the same sub-frame. This is particularly a problem when it is desirable to reduce the number of such reports in a single sub-frame.
In LTE, each sub-frame typically includes two slots, each slot in turn consisting of a plurality of OFDM-symbols. In the frequency domain, each OFDM-symbol can be seen as a set of sub-carriers. The sub-carrier spacing is typically 15 kHz and the number of sub-carriers in the set depends on the bandwidth of the frequency carrier. Further, the sub-carriers are divided into groups of multiple adjacent sub-carriers, e.g. 12 sub-carriers. Each group of sub-carriers in a slot is generally referred to as a “Resource Block”. Within a sub-frame, these resource blocks are arranged as resource block pairs in which information can be conveyed.
As explained above, if a terminal has been scheduled data packets in multiple downlink sub-frames, the terminal is typically required to transmit multiple feedback reports in a single uplink sub-frame. However, single carrier properties must be retained in uplink transmissions according to LTE. As a result, a terminal cannot transmit in more than one resource block and still transmit a single carrier signal, since the corresponding sub-carriers are not in contiguous spectrum, i.e. adjacent frequencies.
Further, if the terminal would transmit multiple feedback reports within a single resource block, the combined signal would typically still not retain its single carrier property as the feedback reports must be transmitted with different CDM (Code Division Multiplexing) code sequences within the resource block, thereby being uncorrelated. In other words, it is typically only possible for a terminal to transmit one feedback report at a time and still maintain the single-carrier properties.
If BPSK (Binary Phase Shift Keying) modulation is used, one bit, i.e. 1 or 0, is conveyed per symbol and the terminal can therefore transmit one feedback report in an uplink sub-frame. By using QPSK (Quadrature Phase Shift Keying) modulation, it is possible for the terminal to convey two feedback reports in an uplink sub-frame, as QPSK allows for two bits per symbol. Even higher modulation schemes, e.g. 16QPSK allowing 4 bits per symbol, are deemed too sensitive to signal disturbances generally resulting in unacceptable error rates. Increasing the modulation order will generally decrease the robustness of the feedback reports, and it is important that the feedback reports are detected correctly with a relatively high probability. The error probability is preferably in the order of 10−3 to 10−4. Therefore, a higher modulation order than QPSK is not an attractive solution to the problem of reporting multiple feedback reports during a single uplink sub-frame.
However, when a transmission allocation of 8 DL:2 UL sub-frames is used, a terminal that has been scheduled in all downlink sub-frames would need to send at least four feedback reports in an uplink sub-frame. Thus, only two possible feedback reports when using QPSK according to the above are clearly not sufficient. Furthermore, a terminal may receive two data blocks, e.g. MAC PDUs (Packet Data Units), in a single downlink sub-frame, each data block requiring a feedback report. In this case, the terminal would need to send twice as many feedback reports in each uplink sub-frame, making the limitation of sending only two feedback reports while retaining single carrier properties even more significant.
A potential consequence of the limitations above could be that it is not possible to transmit data to a single terminal in all downlink sub-frames, which would “artificially” limit the DL capacity by the lack of feedback opportunities.